Thoughts and notes on Uber's latest earnings

On 6th February 2020, Uber announced its Q4 FY2019 earnings. Below are some of the thoughts I had from reading their press release

Take Rate

Uber defines take rate as the result of adjusted net revenue divided by gross bookings. In a layman’s terms, it is the amount of Uber takes from what riders pay for rides, after paying drivers their share. In Q4 2019, the take rate reached 20.6% compared to 18.7% in Q4 2018. It meant that out of $100 taken from riders, Uber took in more money in 2019 than in 2018

Source: Uber

However, if we look at 2019 as a whole, take rate dropped to 19.8%, compared to 20.7% in 2018, almost a full 1% lower.

Source: Uber

Overall, in the second half of FY 2019, Uber had higher take rates overall, for Rides and for Eats individually than in the second half of FY 2018. However, the gain was sufficient to make up for the deficit of the first half of FY 2019 to the first half of FY 2018. As you can see from the graph above, Eats provided a terrible take rate, compared to Rides.

Driver Incentives and Driver Referrals

The incentives and referrals help reflect the health of the brand and business. Low payout for incentives and referrals means that Uber spent less money to recruit drivers and increase rides. Incentives and referrals are usually presented in this manner by Uber

Source: Uber

I calculated the ratio of Adjusted Net Revenue (ANR) over revenue in 2018 and 2019 for both Rides and Eats. The higher the figures, the better for Uber

Apparently, it keeps getting better and better for Rides. On the Eats side, Uber seemed to recover from the slump in Q3 and Q4 2018.

Rides

Rides continues to be the silver lining in the EBITDA area for Uber. It is the only segment with positive EBITDA in Q4 2019 or FY 2019 as a whole.

Source: Uber

It’s even better for Uber that YoY growth for Rides EBITDA (34%) is bigger than that for Rides Bookings (18%), Revenue (27%) and ANR (30%).

Eats

Eats registered the biggest loss among Uber’s segments in Q4 2019. Uber may find it encourage the fact that Eats’ Q4 loss is only 111% of ANR, compared to 168% in the same period last year.

Uber recently announced the divestiture of Uber Eats in India. Since Uber Eats was losing money and users in India, the decision looked a wise one and in line with the strategy pursued by the company.

Source: Atom Finance

CEO of Uber revealed on the earnings call that Uber Eats in the US made up almost 39% in gross bookings of the global Eats GB ($1.7 bn out of $4.374 bn). There are 400,000 active restaurants in the US on the Eats side, up by 78% YoY.

Freight

Freight’s Q4 loss was a tad more than 25% of its ANR, compared to a bit more than 18% of the same period last year. Not a trend that Uber would want in their quest to become profitable.

Cost structure

On a full year basis, only depreciation as % of revenue decreased in 2019, compared to 2018. Overall, operating cost and expenses increased significantly in 2019, reaching 161% of revenue in 2019. However, Q4 2019 provided a brighter picture for Uber. Only R&D as % of revenue went up in the quarter, compared to the same period last year, especially given that operating expenses as % of revenue in Q4 2018 were higher than those of FY2018.

Good bits of information here and there

  • Uber for Business’ Gross Bookings made up 9% of the total GB
  • In Q4 2019, Uber Rewards Program had 25 million subscribers from multiple markets, up from 18 million from the US alone reported in Q3 2019
  • Multiple-app users had almost 3 times the number of transactions as single users

Thoughts

Though challenges remain, including those posed by local authorities threatening to impose infavorable regulations, driver/rider safety and competition, Q4 2019 seemed to offer the team at Uber and bull investors something to be optimistic about.

In an ideal world, I would love to see more transparency regarding:

  • Margin of products such as Uber for Business, Airport, Helicopter, Comfort, Scooter
  • Margin of Eats in the US or products in the key market
  • More details on subscriptions
  • Engagement data regarding the use of multiple apps per user

Notes from Disney's Q1 2020 earnings call

Today, Disney announced its Q1 2020 results. There are a lot to unpack as the business is pretty diverse. I am just covering some of the stuff I mainly care about.

Overall, revenue increased 36% year over year. The effect from investments in Disney+ is reflected on operating income which increased only by 9% compared to last year

Parks made up 35% of Disney’s revenue, but more than 58% of its operating income. Parks also provided the largest margin at 32% among Disney’s segments, followed by Media Networks.

Subscribers

  • Disney+: 28.6 million paid subscribers as of 3rd February 2020 from US, Canada, Netherlands, Australia and New Zealand
  • ESPN+ 7.6 million paid subscribers as of 3rd February 2020
  • Hulu has 30.7 million paid subscribers as of 3rd February 2020

Given that Disney publicly set a target of 60-90 million paid subscribers worldwide and of profitability in 2024, it is a promising start to reach the 28-million mark already just a few months after launch. Bob Iger wisely tried to play down any enthusiasm from the figures by citing the inability to point out the reason for the growth and uncertainty in the key international markets where Disney+ will debut soon.

Average Revenue Per User

The dip in ESPN+ and Hulu SVOD APRU was attributed to the bundle that offers Disney+, ESPN+ and Hulu Ad Supported for $12.99/month. Regarding the Hulu APRU, it’s even higher the non-ads subscription of $11.99/month. Christine McCarthy, Disney’s CFO, had the following comment:

The ad supported, the product is priced at $5.99. And but the ad-supported part of the equation makes the ARPU come out even higher than the ad-free. Most of the subscribers subscribe to the ad-supported. So that’s a good balance of the ARPUs when you stack them up next to each other.

Source: Atom Finance

Regarding the APRU of Disney+, since the service is offered at different pricing tiers including the promotion with Verizon, the 3-year plan last year, the bundle and full price, it’s difficult as to what to make out of the figure. Below are a few things from the earnings call:

  • 50% subscribers came directly from disney.com
  • Bob Iger mentioned “20% of those subscribers” came from Verizon. The comment in the earnings call wasn’t clear, but he clarified it in this interview with CNBC
  • Most subscribers came from the US
  • Conversion from free-to-pay and churn rates were better than what Disney had expected
  • No significant churn after Mandalorian Season 1 ended

Other notes

  • “It was 65% of the people who watch Mandalorian watch at least 10 other things”
  • Each Disney+ subscriber spent 6-7 hours every week on the service
  • 18-22% guests to parks were international guests
  • “Attendance at our domestic parks was up 2% in the first quarter, and per capita guest spending was up 10% on higher admissions, merchandise and food and beverage spending. Per room spending at our domestic hotels was up 4%, and occupancy was 92%. So far this quarter, domestic resort reservations are pacing up 4% compared to this time last year, and booked rates at our domestic hotels are currently pacing up 10%.”
  • The fight between McGregor and Tyrone brough “1 million pay-per-view purchases and 0.5 million new subscribers”

Disclosure: I own Disney stocks in my personal portfolio

Netflix's change in view accounting and a misleading use of Google Trends

This week, Netflix dropped its latest earnings report. There are a lot of positive announcements from Netflix and kudos to them for weathering the rising competition from a plethora of streamers, so far. Nonetheless, there are a couple of notable points that I am either intrigued by or in disagreement with.

How a view is counted

Netflix used to register a view whenever a user passed a 70% of a show or a movie. Recently, the company changed that policy. According to the latest earnings report, whenever a viewer reaches a two-minute mark, it counts a view.

Source: Netflix

Netflix communicated the change in a tricky and inconspicuous manner. The explanation on the two-minute mark only came in the footer; which certainly isn’t where readers’ attention focuses on.

Source: Netflix

As you can see in the last sentence in the screenshot above, the change in the view accounting usually results in an increase in view because of obvious reasons. I don’t believe two minutes is enough to determine the intention of the audience. It is not uncommon that viewers watch 20 minutes of a show or a movie before leaving. If Netflix thinks that 70% is too high a standard, 40% or 50% would make more sense than the new implemented policy.

Using Google Trends to compare The Witcher with Mandalorian

In the Competition section of the report, Netflix dropped a Google Trends screenshot that showed interest in its currently flagship show The Witcher, Mandalorian, Jack Ryan and The Morning Show in the last 90 days worldwide

Source: Netflix

First of all, I am not sure this is an apple-to-apple comparison due to the difference in availability. Disney Plus is only available in US, CA, Australia, New Zealand and Netherlands. Even though Prime Video is supposedly accessible worldwide, while I was in Vietnam, I couldn’t watch many shows on the platform despite my membership.

Netflix said that even if Disney+ were global, the results wouldn’t be much different, citing the following result on Google Trends

I wouldn’t make that claim with such a degree of certainty. Netherlands is just a small country in Europe with about 17 millions in population. The viewership and interest in that country doesn’t equal to those worldwide.

Furthermore, the shakiness of the comparison can also come from the selection of keywords. Since The Witcher or Witcher is the name of a video game released in 2017, neither of the two keywords isn’t exclusive to the show on Netflix. Unfortunately, Google Trends doesn’t offer a feature that can clearly separate the show and the video game. The best that we can do is to filter the results by categories. I tested it out by comparing the keywords: Witcher, The Witcher,

As the screenshot shows, there is a big different between “Witcher” and “The Witcher”. The gap is even starker when “Netflix” is added to the search terms. If we set “Art and Entertainment” as the category, the picture will look a bit different

The Witcher/Witcher keywords had a spike on 21st December 2019, one day after the launch of the Netflix show while Mandalorian hit its peak on 28th December 2019, one day after the season finale. The difference between the yellow line and the red line is closer when we look at “Art & Entertainment” alone than when we look at “All categories” which may likely include the effect from The Witcher video game.

Now, the result above still doesn’t offer the full picture thanks to the difference in geographical availability, Let’s look at the US, two markets where every show is available

If we look at the United States alone for All Categories, it looks more favorable for Mandalorian. When “Art & Entertainment” is applied, it fares even better for Mandalorian

Here are the results for Canada

What about Australia?

My point is that there are several factors that affect how the search terms are presented on Google Trends and how results should be interpreted. I don’t have an idea on how the competing shows actually fare. I do believe that the way Netflix presented the information and data in its report is misleading at best.

Disclosure: I have Disney and Apple stocks in my portfolio.

Baggage fees by airlines for domestic flights in the US

I came across this piece of news from Skift

JetBlue Airways is raising fees for checking bags again by $5 — to $35 for the first one and $40 for the second — on flights within the United States.

Baggage fees in the US have become a significant source of revenue for airlines in the US. In 2018, baggage fees brought more than $4.5 billion in revenue for major airlines in the US, compared to $1.1 billion in 2008. Throughout the first three quarters of 2019, the figures are well on their way to surpass the 2018 ones. Though I understand the monetary perspective through which many look at this issue, I find it annoying that airlines seemingly take advantage of customers this way. We often travel with luggage and for some certain routes, there are very few options as only one or two carriers operate on the routes. Customers have little freedom to choose.

I compiled the baggage policy of major US airlines below, as well as some information on their baggage revenue and how much of the total revenue it makes up.

Source: Bureau of Transportation and official airlines’ websites

“Low price” airlines such as Frontier and Spirit have a significant portion of their revenue from baggage fees. The low prices are often misleading as we rarely travel without a carry on. Only an addition of carry-on fees will reveal the whole cost of a flight ticket with the low price airlines. Among the bigger carriers, as far as I know, only United Airlines (not shocked at all) charges customers for carry-on in certain cases.

Which business segment is the most profitable for McDonald's?

Overall business

McDonald’s main business segments include the company operated and franchised outlets. At the moment, franchised restaurants make up 93% of the total store count, not so far off the company target of 95%. Under the franchise segment, there are conventional franchise, developmental and foreign affiliate agreements, each of which comes with a different revenue and expense structure. Below is a diagram I quickly drew to summarize McDonald’s business structure

Getting into the weed

Between the company-operated restaurants and franchised counterparts, the latter have a much higher margin

Source: McDonald’s

The franchised outlets posted a 82% gross margin in 2018, compared to just about 17% by the company-operated restaurants. The three-year figures also indicated the shift of focus on being a franchisor. Revenue from franchised restaurants grew on average by 8% every year from 2016 to 2018. In the same period, the domestically run outlets recorded a decline of around 19% on average every year.

The focus on being a franchisor is also reflected on the lease agreements that the company has

Source: McDonald’s

In 2018, the company was on the hook for smaller rent expense at company-operated restaurants than at franchised outlets.

McDonald’s defines its main geographical segments as follows:

  • U.S. – the Company’s largest segment.
  • International Lead Markets – established markets including Australia, Canada, France, Germany, the U.K. and related markets.
  • High Growth Markets – markets that the Company believes have relatively higher restaurant expansion and franchising potential including China, Italy, Korea, the Netherlands, Poland, Russia, Spain, Switzerland and related markets.
  • Foundational Markets & Corporate – the remaining markets in the McDonald’s system, most of which operate under a largely franchised model. Corporate activities are also reported within this segment.
Source: McDonald’s

Systemwide revenue has been decreasing since 2016. The main reason for the decline was attributed to company-operated segment as it shrank by 21% and 17% year over year in 2018 and 2017 respectively. Additionally, the chart showed that International Lead Market was the only segment that registered revenue growth in 2018. High Growth Market delivered the biggest growth in franchised revenue, but also the biggest decline in company-operated revenue. Concerningly, the high growth market segment registered a decrease in revenue while it should have been the opposite, essentially due to the decline in company-operated revenue.

In terms of operating margin, Foundational Markets led the way in the franchise department while International Lead Markets took the top spot in the company-operated one.

Source: McDonald’s
Source: McDonald’s

Regarding restaurant counts, systemwide restaurant count went up compared to 2017, but presence in the US shrank.

Source: McDonald’s

It’s helpful when the company gave more color on the company-operated/franchise breakdown of the store count.

Approximately 93% of the restaurants at year-end 2018 were franchised, including 95% in the U.S., 88% in International Lead Markets, 83% in High Growth Markets and 98% in Foundational Markets.

Source: McDonald’s

It is very interesting that the net restaurant addition in 2018 was bigger than that in 2017, but the company added smaller initial fees in 2018 and bigger revenue from rent and royalties than in 2017.

Source: McDonald’s

The most likely explanation I could come up with is that the company gave quite a haircut on the initial fees for new restaurants in High Growth Markets (see the breakdown a few paragraphs down).

The timing of new and closed restaurants varied from one market to another and it made the calculation of average revenur or sale per restaurant tricky. Nonetheless, if we can make simple assumptions and take the revenue as well as restaurant count at year end, we can compare 2018 vs 2017 to see where the business is going

Regarding the US, compared to 2017, average sale and revenue per company-operated restaurant decreased in 2018 while both average revenue and sale per franchised restaurant increased.

For International Lead Markets, compared to 2017, average revenue and sale per restaurant increased for both company-operated and franchised outlets.

The situation in High Growth Markets is similar to that in the US. Average revenue and sale per company-operated restaurant decreased significantly.

Foundational Markets mirrored what happened with High Growth Markets.

In terms of ownership type, conventional franchise is the biggest source of revenue, followed by developmental licensed and foreign affiliate. However, foreign affiliate grew the most in 2018 compared to 2017, by 6.5%, followed by developmental licensed with 4% and conventional franchise with 1.5%.

Source: McDonald’s

When looking at the expense structure for company-operated restaurants, food and paper is the biggest piece of the pie, even though personnel registered the biggest growth YoY compared to 2017.

Above is just my analysis for the operating side of McDonald’s based on its last two annual reports. There is a lot more to look at such as the cash flow, the financing and investing activities. I hope that you found something useful in this entry.

Uber incorporated transit ticketing into its platform

Uber rolled out the feature first in Denver last summer, allowing the city residents to buy public transit tickets straight from its platform. Apparently, it has been a positive experiment so far. The company reported an average growth of 42% each week from May to the end of June 2019. In an interview with Bloomberg Technology today, the Head of Uber Transit said that the feature’s user growth in Denver was 15% week over week. Judging by the difference of the two growth figures, I guess that the latter was taken over a longer time period and a bigger base. Furthermore, Uber announced today that it would bring the ticketing feature to Las Vegas.

Uber also said that the number of repeat ticket purchases has increased every week since ticketing launched. As of the week of June 24th, approximately 25 percent of tickets sold were purchased by users who had previously purchased tickets on the app.

Source: The Verge

I think adding the ticketing option to the platform makes sense.

In the beginning, Uber under Kalanick tried to grow exponentially by reaching as many domestic and international markets as possible. After the change in the leadership, Uber scaled back its presence overseas and sold unprofitable businesses in South East Asia and China. If it couldn’t expand geographically, how growth could be obtained? By going vertically and more deeply in the existing market.

Hence, Uber offers additionally services like Uber Eats, Uber Bike, Uber Freight or Uber Fly. Uber’s transit ticketing is another way to move towards the goal of being a one-stop shop for transportation.

I used to rely on public transit in Omaha a lot. The local bus operator’s website offers a detailed schedule of what time buses approximately will come and leave. However, it doesn’t provide an estimate arrival time and the user interface pitifully leaves a lot to be desired. Map applications such as Apple Maps and Google Maps are fairly helpful in tracking bus movements and giving an estimated arrival time. But it doesn’t provide a comparison between options such as public transit and an Uber ride. Uber’s value propositions lie in its household name, a comparison between options in terms of time and cost, cashless payments and convenience. If you can help users eliminate one click or action on the phone during a user journey, it can be an added value.

Source: Uber

It is unclear whether and how Uber has so far managed to earn revenue and profit from the new feature, though the Head of Uber Transit confirmed that it was a revenue-generating vehicle. Even if there is no money to be made yet, when the application attracts more usage, users and traffic, it will find ways to make more money later on.

Another benefit that I suspect the new feature will bring to the table is to act as an anchor option for its rides. An anchor offer is one that is presented to make the truly primary offer more attractive. It’s similar to readers, after seeing a package of online and offline access to a newspaper have the same price as the exclusive digital access, choosing the combination package. As you can see in the screenshot above, even though the transit option comes with a saving of $8, it will cost a rider 20 minutes more. In some situations, riders may be more motivated to choose an Uber ride, instead of waiting for a public transit.

Uber may not make any money when it sells bus and subway tickets through its app, but it is seeing an uptick in business as a result. Since Uber launched its transit planning feature in January, Uber trips in Denver that start or end at a transit station have grown 11.6 percent. This helps bolster Uber’s claim that it is helping solve the first mile / last mile challenge that plagues many cities.

Source: The Verge

So the transit option is just one way in which we are increasing our relevance to a greater number of consumers on a global basis. And we are seeing it in higher engagement in the app specifically with London and some of the other areas where we’ve grown transit.

Source: Uber’s Q2 Earnings Call Transcript

Furthermore, The Verge reported a concerted effort by Uber to appear less contentious towards public transit. Working with various stakeholders in the markets in which they operate will earn Uber some goodwill. For a business endlessly engaged in legal issues with local authorities across the world, some goodwill is definitely helpful.

Starbucks in Vietnam

I came across an interesting video by CNBC on Starbucks’ alleged struggle in Vietnam.

I agree with several points brought up in the video clip, but there isn’t much data to back up their thesis that Starbucks is struggling in Vietnam. The only corroborating data is a Starbucks store per capita. Apparently, the figure in Vietnam is much lower compared to that in our neighboring countries.

Nonetheless, there is no financial data that indicates Starbucks is struggling in our country. Moreover, Starbucks has been expanding slowly yet steadily in Vietnam.

To show that Starbucks is struggling, the video reported that the chain owns less than 3% of the market. Well, which market? If you’re talking about the market that involves every mom-and-pop shop in every province and everybody across all income brackets, then I am not surprised at the low market share. Coffee at street stalls is significantly cheaper than coffee at Starbucks and I doubt you can find the chain outside of the major cities such as Hanoi, Saigon or Danang. But Starbucks isn’t for everyone. To make the market share figure more relevant, it has to be for the upscale market. Compare Starbucks with Runam, the Workshop and stores in shopping malls, and see how the green brand fares. My experience with Starbucks in Saigon is that it never lacks traffic. It always seems popular to the citizens. The point is that the video lacks quantitative evidence to make the case.

There is one more example that I want to add regarding how Starbucks doesn’t really fit in the Vietnamese culture. We Vietnamese people like to go to coffee shops and spend hours there working or browsing the Internet. In Starbucks stores in Vietnam, you are given one Wifi code for each receipt and each code is good for two hours, I believe. Asking for one more code is possible, but if a customer forgets, it will be inconvenient for an average Vietnamese customer.